<正> Objective: To study the clinical relations of portal hy-pertensive gastropathy (PHG) of hepatitis B cirrho-sis to other factors.Methods: Three groups of subjects were studied pro-spectively at our hospital from March 2000 to March2001: 159 hepatitis B cirrhotic patients with portalhypertension, 114 hepatitis B cirrhotic patients with-out portal hypertension, and 97 control subjects.Free portal vein pressure (FPP) was measured dur-ing surgery. Liver function was assessed by Pugh'smodification of Child's criteria. The area of livercollagen fibrin was studied using color image analysissystem. Esophageal varices were identified by Dagra-di grading. Gastric varices were identified accordingto Northern Italian Endoscopic Council (NIEC)grading. Hypersplenism was assessed with the reduc-tion of WBC, HGB and PLT. Hepatitis B virus inthe gastric mucosa was detected by immunizing histo-chemistry. Helicobacter pylori (H. pylori) organismswere identified by rapid urease testing and/or exami-nation of the stained biopsy specimens (haematoxylinand eosin). To analyze the correlation between theseendoscopic signs at the gastric level and other fac-tors.Results: The differences of FPP among the threegroups (patients with grade Ⅰ, Ⅱ, and Ⅲ gastropa-thy) were not significant. There was no correlationbetween Child-Pugh classification grading and the se-verity of gastropathy (P=0. 153). The differences ofthe area of liver collagen fibrin among the threegrade gastropathy were not statistically significant (P=0. 801). There was a significant difference in theprevalence of severe PHG among grade Ⅰ, Ⅱ, Ⅲ,Ⅳ and Ⅴ esophageal varices (P<0. 001). PHG waspresent in a similar percentage of patients with gas-tric varices compared with those without gastric vari-ces (P=0. 209). There was a significant difference inthe severity between PHG and hypersplenism (P=0. 003). Seven patients with PHG had no microscopicevidence of hepatitis B virus infection in the gastricwall. There was no correlation between Child-Pughclassification grading and infection of H. pylori (P=0. 7491).Conclusions: The most important element causingPHG is the increased portal pressure as a prerequi-site. In addition, other factors may contribute to thedevelopment of PHG. PHG often occurs in patientswith the presence of esophageal varices. There is amarked correlation between the severity of PHG andhypersplenism. Hepatitis B virus and H. pylori infec-tion are unlikely to be involved in the pathogenesis ofPHG. The development of PHG is less influenced ei-ther by the severity of liver disease (Child-Pughgrade) and cirrhosis or by the presence or non pre-sence of gastric varices.
